Political Quotient (PQ)

Kay Chandler: You're not a Democrat, are you?
Dutch Van Den Broeck: What if I am?
Kay Chandler: We talk. I give you books to read
--Random Hearts

Central to the method for measuring media bias in Groseclose (2011) is the political quotient (PQ). The PQ is a number, generally between 0 and 100, that reflects how liberal an individual is. The higher the number, the more liberal.

The PQ estimates the extent to which an individual agrees with endorsed positions of Americans for Democratic Action (ADA), a liberal interest group. Thus, to determine liberal media bias, the PQ employs criteria selected by liberals.

This is an attractive approach, because it reduces the contestability of what constitutes the liberalism construct. Liberals are defining it themselves. Stated differently, the PQ metric garners considerable validity because it is difficult to contest that the ADA issues are not a legitimate representation of the liberal political agenda.

The PQ of members of Congress can be estimated in a straightforward manner. Whenever an issue supported by the ADA (here is a current list of ADA issues) comes up for vote in the House or Senate, then a Congressperson's liberal bias is expressed by whether his/her roll call vote sides with the ADA's endorsed position.

The PQ is basically the number of times that an individual's roll call vote sides with the ADA position, divided by the total number of ADA-related issues put to a vote. In his method, Groseclose normalizes the scores so that a 50 represents the score of the average member of Congress (here is a list of some well-known politicians and their PQs).

Thus a PQ greater than 50 implies greater agreement with the liberal political agenda than average, or someone who is 'more liberal.'

People who don't happen to be members of Congress can estimate their PQs by assessing how they would have voted on the same issues. Here is a quiz that employs roll call votes taken in the House and Senate in 2009 on ADA-centric issues. Assume some margin for error here, as more accurately pinning down your 'real' PQ may require a larger sample size of roll call issues.

One concern that I have with this approach is that it assumes a unidimensional scale for measuring political bias. Of course, this is a common assumption that facilitates positioning political bias as being on either one side (Left) or the other (Right).

As we've noted on these pages, it is more likely that political beliefs are more appropriately captured on multidimensional rather than on unidimensional scales--which means that the oft referred to 'political spectrum' is more accurately thought of as a 'political landscape.'

Consequently, the PQ scale may commingle groups that have importantly heterogenous beliefs. For example, a libertarian might vote in favor of ADA agenda items related to particular social and foreign policy issues, while voting against ADA agenda items related to economic and fiscal policy issues. On the PQ scale, then, libertarians might appear as moderately liberal.

Groseclose, for example, estimates libertarian Ron Paul's PQ as about 31. However, it is difficult to view Paul's position on the PQ as completely consistent with the well-known politicians bracketing him on Groseclose's example list.

That said, Groseclose's PQ scale offers a worthy building block for estimating media bias. We'll discuss how in future posts.